Logging parameters and metrics in MLOps - Mini Project: Build & Apply

Logging Parameters and Metrics in MLOps

📖 Scenario: You are working on a machine learning project. You want to keep track of the settings (parameters) you use and the results (metrics) you get. This helps you understand which settings work best.

🎯 Goal: Build a simple Python script that logs model parameters and metrics using a dictionary. You will create the data, add a configuration, log the main results, and then display the logged information.

📋 What You'll Learn

Create a dictionary called params with exact keys and values for model parameters

Create a dictionary called metrics with exact keys and values for model evaluation

Add a variable called experiment_name with a specific string value

Log the parameters and metrics into a single dictionary called log

Print the log dictionary to display all logged information

💡 Why This Matters

🌍 Real World

Logging parameters and metrics is essential in machine learning projects to track experiments and improve models over time.

💼 Career

Data scientists and MLOps engineers use logging to monitor model performance and reproduce results reliably.

1

Create model parameters dictionary

Create a dictionary called params with these exact entries: 'learning_rate': 0.01, 'batch_size': 32, 'optimizer': 'adam'.

MLOps

# Create the params dictionary with the given keys and values
# Your code here

Need a hint?

Use curly braces {} to create a dictionary. Separate keys and values with a colon :.

2

Create model metrics dictionary and experiment name

Create a dictionary called metrics with these exact entries: 'accuracy': 0.85, 'loss': 0.35. Also create a variable called experiment_name and set it to the string 'exp_001'.

MLOps

params = {'learning_rate': 0.01, 'batch_size': 32, 'optimizer': 'adam'}
# Create the metrics dictionary with the given keys and values
# Create the experiment_name variable with the given string
# Your code here

Need a hint?

Create another dictionary like params for metrics. Use a simple assignment for experiment_name.

3

Log parameters and metrics into a single dictionary

Create a dictionary called log that contains three keys: 'experiment' with the value of experiment_name, 'parameters' with the value of params, and 'metrics' with the value of metrics.

MLOps

params = {'learning_rate': 0.01, 'batch_size': 32, 'optimizer': 'adam'}
metrics = {'accuracy': 0.85, 'loss': 0.35}
experiment_name = 'exp_001'
# Create the log dictionary combining experiment_name, params, and metrics
# Your code here

Need a hint?

Use a dictionary with keys pointing to the existing variables. For example, 'experiment': experiment_name.

4

Print the logged information

Write a print statement to display the log dictionary.

MLOps

params = {'learning_rate': 0.01, 'batch_size': 32, 'optimizer': 'adam'}
metrics = {'accuracy': 0.85, 'loss': 0.35}
experiment_name = 'exp_001'
log = {'experiment': experiment_name, 'parameters': params, 'metrics': metrics}
# Print the log dictionary
# Your code here

Need a hint?

Use print(log) to show the logged data.